DOP28 Yield of surveillance colonoscopy in patients with Primary Sclerosing Cholangitis

Abstract

Abstract Background Patients with primary sclerosing cholangitis (PSC) have a higher risk of developing colorectal cancer (CRC), in particular in patients with concurrent diagnosis of inflammatory bowel disease (IBD). Therefore, strict surveillance colonoscopy in patients with PSC and PSC-IBD is recommended with consideration of random biopsies. The aim of this study was to assess the method and yield of surveillance colonoscopy in patients with PSC and PSC-IBD in routine practice in the Netherlands. Methods An observational cohort study was performed within the EpiPSC2 cohort in The Netherlands. Patients with a confirmed PSC diagnosis were included from January 2008 onwards. Data were collected retrospectively from the date of diagnosis until inclusion, and prospectively from that date forward. Patient demographics, PSC and IBD disease characteristics were collected at study inclusion and during annual follow-up. For this study, data collection was expanded retrospectively with colonoscopy and pathology reports. Outcomes were endoscopic surveillance technique, whether random or targeted biopsies were taken, if polypectomy was performed, dysplasia detection rate and type of dysplasia. Results The study population included 1,203 PSC patients, of whom 82.1% had large duct PSC, and 70.6% concomitant IBD (table 1). A total of 4,099 colonoscopies were performed in 759 individual patients, with a median of 4 colonoscopies per patient (IQR 2 – 8). The median follow-up duration was 11 years (IQR 12 – 17). In total, 11.5% (n=473) of colonoscopies were performed with chromo-endoscopy. Among the conducted colonoscopies, 2,543 included biopsies, 355 involved polypectomies, and 86 comprised both biopsies and polypectomies. Dysplasia was identified in 11.4% (341 out of 2,984) of the colonoscopies, including 14.1% indefinite for dysplasia (n=48), 75.1% low grade dysplasia (n=256), 7.6% high grade dysplasia (n=26), and 5.9% adenocarcinoma (n=20). The detection of dysplasia occurred through random biopsies in 89 colonoscopies (26.1%), targeted biopsies in 114 colonoscopies (33.4%) and polypectomy in 153 colonoscopies (44.9%) (figure 1). Dysplasia was observed in 195 (26%) individual patients, of whom 80.5% had PSC-IBD. Conclusion Our data highlights the importance of colonoscopic surveillance in PSC and PSC-IBD patients. Dysplasia was detected in over a quarter of patients undergoing surveillance colonoscopy. Notably, 26.1% of dysplasia cases were identified through random biopsies, emphasizing the importance of incorporating random biopsies into surveillance protocols for this specific patient population.